Introduction

Lantibiotics are ribosomally synthesised and post-translationally modified antimicrobial peptides produced by bacteria. This class of bacteriocins contains atypical amino acids, such as dehydroalanine (Dha), dehydrobutyrine (Dhb), lanthionine (Lan) and β-methyllanthionine (MeLan) residues (Chatterjee et al. 2005; Field et al. 2010). The lantibiotic nisin is used as a safe and natural preservative in the food industry (Delves-Broughton et al. 1996). The biosynthesis of nisin requires the products of 11 biosynthesis genes. Prenisin, the inactive translation product of the nisA gene, undergoes several post-translational modifications, of which the key reaction is the selective dehydration of serine and threonine residues in the precursor molecule, leading to Dha and Dhb, respectively. NisB, which is a membrane-associated enzyme, catalyses this dehydration reaction (Chatterjee et al. 2005; Field et al. 2010; Siegers et al. 1996). Some of these modified amino acids form Lan and MeLan residues as a result of the addition of neighbouring cysteine thiol groups to the unsaturated side groups, catalysed by the cyclase NisC (Li et al. 2006).

Nisin belongs to the type-A(I) subgroup of lantibiotics that contains other well-characterised linear lantibiotics, including subtilin, epidermin and Pep5 (Chatterjee et al. 2005). In these groups of peptides, the Lan and MeLan residues are formed by the action of LanB and LanC proteins respectively. LanB proteins have no close sequence similarity with known proteins in the databases other than members of the LanB family. LanC proteins have a close sequence similarity with members of the LanC family and with the C terminus of proteins that belong to the LanM family (Chatterjee et al. 2005; Siezen et al. 1996). The LanC-like proteins (LANCL) have 20–25% sequence identity with the NisC protein. These proteins are found in a wide range of organisms that do not produce lantibiotics, such as Drosophila melanogaster, Xenopus laevis, Arabidopsis thaliana, and some mammals (Li et al. 2006). NisB shares the closest homology (42% identical residues) with the protein NsuB from Streptococcus uberis, a natural nisin variant nisin U producer organism (Wirawan et al. 2006).

NisB dehydrates the serine and threonine residues in the precursor peptide (Chatterjee et al. 2005; Karakas Sen et al. 1999). However, no experimental evidence for the in vitro dehydration function of NisB protein has been reported to date (Field et al. 2010). The structural genes, the nisB gene and other genes responsible for the biosynthesis of nisin have been sequenced (Dodd et al. 1990; Kuipers et al. 1993). NisB is a large membrane-associated protein of 993 amino acid residues with a calculated molecular weight of 117.5 kDa (Engelke et al. 1992). Its sequence contains a number of potential amphipathic membrane helices, and it has been proposed that nisin modification enzyme NisB forms a membrane-associated complex with the cyclase NisC and the transporter NisT (Siegers et al. 1996; van den Berg van Saparoea et al. 2008).

The purpose of the study reported here was to employ recombinant methods to produce substantial amounts of NisB which can later be used for the in vitro investigation of the dehydration function of NisB protein. In past studies, some difficulties were encountered by ourselves as well as by others in attempts to isolate a potentially unstable membrane-bound complex (Karakas Sen et al. 1999; Kupke and Gotz 1996). In this study, we have used a His-tagged approach for the purification of NisB protein. Here we report the heterologous expression of the nisB gene of a nisin A-producing Lactococcus lactis strain in Escherichia coli and its partial purification.

Materials and methods

Microbiological techniques, strains and plasmids

The L. lactis and E. coli strains used in this study are listed in Table 1. L. lactis FI5876 (Dodd et al. 1990) was grown in M17 medium (Oxoid, Basingstoke Hampshire, UK) supplemented with 0.5% (wt/vol) glucose (GM17 medium) at 30°C without agitation. E. coli strains were routinely grown in L broth at 37°C on an orbital shaker. Media plates were made by the addition of 1.5% (wt/vol) agar to liquid broth. Antibiotics (Sigma-Aldrich, St. Louis, MO) were added as selective agents when appropriate: ampicillin 50 μg ml-1 (for selection in broth) and 200 μg ml-1 (for selection of transformants on plates); chloramphenicol 34 μg ml-1. The more stable carbenicillin (50 μg ml-1) was used in place of ampicillin in cultures grown for protein purification. The chromogenic substrate 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-Gal) was used at a final concentration of 40 μg ml-1. The E. coli cells were induced with 1 mM isopropyl β-D-1-thiogalactoside (IPTG) for recombinant protein expression.

Table 1 Strains and plasmids used in this study
Full size table

Molecular techniques

Plasmid DNA isolation, DNA manipulation and transformation of E. coli strains with recombinant plasmids were carried out as previously described (Dodd et al. 1990). Restriction enzymes and other DNA-modifying enzymes from various sources were used according to the suppliers’ recommendations. Restriction DNA fragments were isolated from an agarose gel by electrophoresis on a DEAE-cellulose membrane using the method described by Dretzen et al. (1981). Caesium chloride (CsCl) gradient DNAs of pFI2268 and pFI381 (Table 1) were prepared according to Sambrook et al. (1989). Primers were synthesised on an Applied Biosystems DNA synthesiser (model 381A; Applied Biosystems, Foster City, CA). The PCR was performed with the following thermal cycling programme: an initial cycle of 2 min at 96°C, followed by 5 cycles of 1 min at 92°C, 1 min at 37°C and 1 min at 72°C, then by 20 cycles of 1 min at 92°C, 1 min at 55°C, and 1 min at 72°C and terminated with a final cycle of 1 min at 92°C, 1 min at 55°C and 10 min at 72°C. Fragments generated for the construction of vectors were amplified with DyNAzyme λ DNA polymerase (Flowgen Biosciences, Nottingham, UK) and cloned into pCR 2.1-TOPO vector (Invitrogen, Carlsbad, CA) before nucleotide sequence confirmation. For routine PCR screening of recombinant clones, AmpliTaq DNA polymerase (Perkin-Elmer, Foster City, CA) was used. The nucleotide sequences of PCR-generated fragments were confirmed on purified plasmid DNA with an Applied Biosystems DNA sequencer (model 373A) and the manufacturer's Taq DyeDeoxy Terminator Cycle Sequencing kit.

Cloning of the nisB gene into expression vector pET-15b

The N-terminal region of the nisB gene was amplified by PCR from chromosomal DNA of L. lactis FI5876 (Fig. 1A, a) using primers NisB-Fwd and NisB-Rev (Table 2). Primer NisB-Fwd included an NdeI site which is in frame with the start codon of the nisB gene. Primer NisB-Rev was close to the downstream site of MunI. An amplified 332-bp fragment was initially ligated into T/A cloning vector pCR 2.1-TOPO using the TOPO TA cloning kit (Invitrogen), and E. coli TOP10 competent cells were transformed with this ligation mixture according to the manufacturer’s instructions. Transformants were selected on L-agar plates containing ampicillin and X-gal and were screened by PCR using M13 Reverse and Universal M13 Forward primers (Table 2) for the presence of recombinant plasmids. Plasmid DNAs were prepared from the positive clones and analysed by restriction digestion using EcoRI for confirmation of the expected construct. The identity of the inserted fragment was further confirmed by DNA sequencing. This plasmid was labelled as pFI2268 (Fig. 1A, b), and the E. coli strain harbouring pFI2268 was designated as FI9748 (Table 1).

Table 2 Oligonucleotide primers used in this study
Full size table

Plasmids pFI2268 and pFI381 were purified by CsCl gradient centrifugation. The BamHI/MunI fragment containing the 5′ end of the nisB gene was extracted from plasmid pFI2268 and ligated into the equivalent BglII/MunI sites of plasmid pFI381 to regenerate an intact nisB gene (Fig. 1A, c, d). E. coli TOP10 cells were transformed with this ligation mixture, and transformants were selected on ampicillin plates. Some transformants were screened by colony PCR using M13 Forward and Reverse Universal primers to confirm the presence of recombinant plasmid. In addition, plasmid DNA was prepared from transformants and analysed by restriction digestion. A plasmid containing the correct fragment was selected and labelled pFI2269 (Table 1; Fig. 1A, d). The E. coli strain harbouring this plasmid was designated as FI9751 (Table 1).

The plasmid pFI2269 was digested with NdeI and BglI, which generated a 3.0-kb NdeI/NdeI fragment containing the intact nisB gene. BglI was included in the digestion as it cuts the vector into small fragments, facilitating the isolation of the nisB fragment from the agarose gel. The NdeI/NdeI fragment was ligated with NdeI-digested pET-15b (Fig. 1A, e, f) and the ligation mixture was used to transform E. coli TOP10 competent cells. Transformed colonies were screened by PCR using primers T7 Promoter-Fwd and NisB-Rev (Table 2), and further confirmation was provided by NdeI digestion of plasmid DNA. The orientation of the inserted fragment was checked by sequencing from both ends of the inserted segment using primers T7 Promoter-Fwd and T7 Terminator-Rev. The resulting plasmid, pET-15b-nisB, was named pFI2270 (Table 1; Fig. 1A, f), and the competent E. coli host strain with pFI2270 was designated FI9752.

Expression of the nisB gene in E. coli

The pET-15b construct (pFI2270) carrying the nisB gene encoding His6-tagged NisB (His6-NisB) was used to transform the expression hosts E. coli BL21-Gold(DE3)pLysS and E. coli BL21 (DE3), generating FI9753 and FI9755, respectively. Transformation of these E. coli hosts with pET-15b yielded the strains FI9754 and FI9756, respectively (Table 1). E. coli cells were grown at 37°C to an OD600 of 0.6–0.8 with the appropriate antibiotic (Table 1). His6-NisB expression was induced by adding IPTG and cultivation was then continued at 37°C for 4 h. All subsequent steps were performed at 4°C. Cells were harvested by centrifugation, and the pellet was washed once with 200 mM Tris-HCl, pH 8.0, and resuspended in the same buffer containing 1 mM dithiothreitol (DTT) and 1 mM phenylmethylsulfonyl fluoride (PMSF). The cell suspension was sonicated at 20 amplitude microns using an MSE Soniprep (MSE, London, UK). To prevent heating, the vessel containing the cell suspension was surrounded by ice. Sonication was applied in short bursts of 20 s (with 1-min cooling period) with a total exposure time of 1.7 minutes. The sonicate was centrifuged at 12,000 g for 30 minutes at 4°C. The supernatant was designated as the soluble fraction. The pellet was resuspended in the membrane extraction buffer [100 mM Tris-HCl pH 7.0, 8 M urea, 3% sodium dodecyl sulfate (SDS), 0.5% Triton X-100] containing 1 mM DTT and 1 mM PMSF, vortexed vigorously, and centrifuged for 2 min at 3,000 g. The resulting supernatant was designated as the membrane fraction. The fractions were analysed by SDS-polyacrylamide gel electrophoresis (PAGE) and western blot analysis as described by Karakas Sen et al. (1999).

Purification of His6-NisB protein

For purification of the His6-NisB protein, the cultures of FI9753 were grown at 37°C to an OD600 of 0.6–0.8 in L Broth containing 50 μg ml-1 carbenicillin and 34 μg ml-1 chloramphenicol. Expression of the His6-NisB protein was induced by adding IPTG to a final concentration of 1mM, and cultivation was then continued at 25°C for 5 h. The cells were then harvested by centrifugation. The pellet was washed once with 20 mM phosphate buffer (pH 7.4) and resuspended in the same buffer containing 10 mM imidazole, 500 mM NaCl, 1 mM DTT and 1 mM PMSF. Cells were sonicated as already described. After centrifugation at 20,000 g for 30 min at 4°C, the supernatant was filtered through a 0.45-μm filter and loaded onto a pre-charged nickel affinity column. His6-NisB protein was purified on a HiTrap chelating metal-affinity column (Amersham Pharmacia Biotech, Uppsala, Sweden) on a Fast Protein Liquid Chromatography (FPLC) system (Pharmacia, Uppsala, Sweden), including a Single Path Monitor UV-1 Control Unit for monitoring absorbance at 280 nm. Buffers used to run the Ni2+ affinity column were the binding buffer (20 mM phosphate buffer pH 7.4 containing 10 mM imidazole and 500 mM NaCl) and the elution buffer (20 mM phosphate buffer pH 7.4 containing 1 M imidazole and 500 mM NaCl). After the protein was loaded onto the Ni2+ affinity column, it was washed with the binding buffer. Bound proteins were eluted on an imidazole gradient from 10 mM to 1 M. The collected fractions were analysed by SDS-PAGE for the presence of His6-NisB protein, and His6-NisB was confirmed by western blot analysis using antibodies specific for NisB or the His6-tag (Roche Diagnostics, Indianapolis, NJ), as previously described (Karakas Sen et al. 1999). To remove salt and imidazole, NisB preparation was eluted through a Sephadex G-25 column (Pharmacia, Uppsala, Sweden) and then concentrated by using a Centricon-50 filter (Millipore, Billerica, MA). Protein concentrations were measured by a procedure described by Bradford (1976) using a protein assay kit (Bio-Rad, Hercules, CA) with bovine serum albumin (BSA) as the standard.

Circular dichroism spectroscopy

To quantify the secondary structure of the purified His6-NisB, the protein sample of His6-NisB was subjected to circular dichroism (CD) spectroscopy analysis at room temperature using a J-710 spectropolarimeter (JASCO, Great Dunmow, Essex, UK) in a nitrogen atmosphere. The secondary structure was monitored in the far-UV (195–250 nm) wavelength region using a cell of path length 0.2 mm. The protein concentration was 70 μg ml-1. CD spectra were collected as an average of three successive scans using a computer connected to the spectropolarimeter. The buffer scans were subtracted from the protein scans, and the spectra were then smoothed using software provided by JASCO. The CD data are given as molecular ellipticity [θ] (mdeg). The His6-NisB protein was analysed in 20 mM sodium phosphate buffer (pH 7.4) containing 100 mM NaCl.

Preparation of substrate prenisin

Prenisin-His6 is a 62 amino acid-long peptide that includes the unmodified nisin with a 22-amino acid length leader peptide at the N-terminus and 6-histidine residues added at the C-terminus using the same pET-15b system (Novagen, Madison, WI), similar to that used for the heterologous expression of NisB protein. The peptide was again purified using a nickel affinity column.

Results and discussion

In order to express the NisB protein with six histidine residues on the N-terminus, the nisB gene of L. lactis was cloned into the expression vector pET-15b. The pET expression system (Novagen), originally developed by Rosenberg et al. (1987), is now widely used for the expression of genes from different organisms (Deng et al. 2009; Xie et al. 2002). This system has been designed for the exclusive expression of cloned genes and uses a bacteriophage T7 RNA polymerase-promoter system. The nisB gene was initially cloned using a host that does not contain the T7 RNA polymerase gene in order to eliminate plasmid instability (Table 1). Once the plasmid constructions were established in a non-expression host, they were then transferred into expression hosts containing a chromosomal copy of the T7 RNA polymerase gene (λDE3 lysogen) under lacUV5 control, and the expression of the nisB gene was induced by the addition of IPTG. NisB is a large protein of 993 amino acid residues, encoded by a 2982-bp open reading frame. As there is always the possibility that PCR amplification of this length of sequence may introduce incorrect nucleotide bases, we developed a different strategy which only involved amplification of the 5′-292-bp fragment of the nisB gene. Plasmid pFI381 (Table 1; Fig. 1Ac) carrying the nisB gene provided the remaining 3′ end of the coding region. In order to identify the appropriate digestion sites within plasmid pFI381, we generated a restriction enzyme digestion map of the nisB gene using GCG Software (GCG, Genetics Computer Group, Madison, WI), which revealed that a MunI restriction enzyme cuts the nisB gene 292 bp downstream of ATG start codon. Hence, two primers, NisB-Fwd and NisB-Rev, were designed from the published sequence (Genbank accession number L16226) to amplify the 5′-292-bp fragment of the nisB gene. Using this primer set, we were able to successfully isolate the 5′-292-bp fragment of the nisB gene by PCR amplification from the chromosomal DNA of L. lactis FI5876, and cloned it into the vector pCR (Table 1; Fig. 1A, a, b). The creation of the expression construct pFI2270 is shown in Fig. 1A. The EcoRI enzyme analysis of plasmid pFI2268, which carries the 5′-292-bp fragment of the nisB gene and isolated from E. coli FI9748, is shown in Fig. 1B (lane 3), confirming the presence of the 332-bp PCR product. The NdeI digestion of expression vector pFI2270 (pET15b-NisB) showed the presence of the 3.0-kb intact nisB gene fragment (Fig. 1C, lane 1).

Fig. 1
figure 1

Construction of the expression vector for His6-nisin biosynthesis protein (NisB) protein. A Construction of expression vector pFI2270: a Map of Lactococcus lactis FI5876 chromosome showing the location of the nisB gene and inducible nisA promoter (white arrowhead) in relation to other nis genes, b PCR-amplified NdeI/MunI fragment containing the 5′ end of the nisB gene cloned into the pCR 2.1-TOPO vector, generating plasmid pFI2268, c map of pFI381 containing the nisB gene, digested with Bglll and MunI restriction enzymes, d BamHI/MunI fragment containing the 5′ end of the nisB gene cloned into pFI381, generating plasmid pFI2269, e map of expression vector pET-15b showing the cloning region and T7 promoter (black arrowhead), f NdeI/NdeI fragment containing intact nisB gene cloned into pET-15b, creating expression vector pFI2270. The relevant restriction sites, shown above the maps are NdeI (N), MunI (M), EcoRI (E) and the non-functional Bglll/BamHI hybrid site (B*). B Agarose gel electrophoresis of EcoRI-digested plasmid DNAs. Plasmid DNA was isolated from E. coli transformants and digested with EcoRI to examine the presence of NdeI/MunI fragment containing the nisB gene. Lanes: 1 DNA molecular weight marker (1-kb DNA ladder), 3 plasmid pFI2268, 2, 4–6 other recombinant plasmids, 7 100-bp DNA ladder. Arrow indicates the position of the NdeI/MunI fragment, containing the 5′-292-bp fragment of the nisB gene. C Agarose gel electrophoresis of the nisB expression vector pFI2270 (pET-15b-NisB) and pET-15b digested with NdeI. Lanes: 1 pFI2270. 2 pET-15b, 3 1-kb DNA ladder. Arrow indicates the position of the NdeI/NdeI fragment containing the intact nisB gene

Full size image

Expression and purification of His6-NisB

NisB protein with N-terminal His6-tagged (His6-NisB) was expressed in E. coli BL21-Gold(DE3)pLysS and E. coli BL21 (DE3) strains from the pET-15b-derived construct pFI2270. The nisB gene encoding the His6-NisB protein was cloned under the control of a strong T7 promoter for inducible expression. The strain E. coli BL21-Gold(DE3)pLysS contains plasmid pLysS, a pET-compatible plasmid that produces T7 lysozyme, hence suppressing the expression of target genes prior to induction with IPTG. The expression of the nisB gene in these E. coli host strains was confirmed by western analysis with antibodies specific for the NisB protein or the His6-tag. These results revealed that expression in E. coli BL21-Gold(DE3)pLysS yielded more soluble protein than expression in E. coli BL21 (DE3) (data not shown). Therefore, the soluble protein fraction of E. coli FI9753 strain, E. coli BL21-Gold(DE3)pLysS containing expression construct pFI2270, was used for the purification of His6-NisB. Moreover, after induction with IPTG, the lower temperature (25°C) was used to reduce the occurrence of inclusion bodies (Fink 1998). Similar results were obtained when E. coli BL21 (DE3) strain was used as the expression host for the purification of SpaB (subtilin dehydratase) (Xie et al. 2002).

To obtain further insight into the biosynthesis of lantibiotics and the involvement of novel post-translational modification reactions, the purification and characterisation of the LanB and LanC proteins and in vitro demonstration of modification reactions are necessary. There have been very few reports in the literature of attempts to purify the LanB proteins. In one study, the epiB gene was expressed in Staphylococcus carnosus and the EpiB protein purified by a process that involved three chromatographic steps (Peschel et al. 1996). Xie et al. (2002) recently reported that the spaB gene was overexpressed as His6-tag in E. coli as a cytoplasmic protein with the aid of the GroEL/ES molecular chaperons and purified using Ni2+-affinity chromatography.

In our study, the His6-NisB protein was purified from the soluble protein fraction of E. coli FI9753, in a single step on Ni2+ affinity HiTrap column with FPLC. The collected fractions were analysed by SDS-PAGE for the presence of His6-NisB protein (Fig. 2a). A protein band of approximately 115 kDa was detected, which is accordance with calculated molecular weight of 119.692 kDa for His6-NisB. His6-NisB protein eluted at 240 mM imidazole concentration, as a single peak (Fig. 2b). The purity of the protein (>90%) was determined by SDS-PAGE, and the identity of His6-NisB was confirmed by western blot analysis using antibodies specific for NisB or the His6-tag (data not shown).

Fig. 2
figure 2

Purification of His6-NisB protein. The fractions were analysed by 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (a), and the proteins were detected by Coomassie blue staining. Lanes: M protein molecular weight markers, 1 crude supernatant, 2–6 eluted fractions. b Elution profile of NisB on the Ni2+ affinity column. Arrow indicates the position of His6-NisB protein

Full size image

CD spectra

Circular dichroism spectroscopy is a technique commonly used to study the secondary structure of proteins and peptides in solution (Woody 1996). In this study, the secondary structure of the His6-NisB protein was studied using CD spectroscopy. Figure 3 shows the far-UV CD spectrum (195–250 nm) of His6-NisB, which is characterised by a strong double minima at approximately 209 nm and 222 nm and a stronger maximum at approximately 198 nm; these are characteristic peaks for an α-helical protein (Woody 1996). Fitting data using “SELCON”, an estimate of the secondary structure of His6-NisB was 15.4% α-helix, 35.7% β-sheet, 22.4% turn, and 24.5% random structures.

Fig. 3
figure 3

Circular dichroism (CD) analysis of the His6-NisB protein. The far-UV CD spectrum of His6-NisB was obtained on a JASCO J-710 spectropolarimeter according to the procedure given in the Materials and methods

Full size image

In vitro activity of NisB

Studies on the mechanism of action of the modification enzymes can be best performed in vitro. In our study, enzyme assays were carried out using either crude extracts of the NisB from L. lactis strain overexpressing the nisB gene (Karakas Sen et al. 1999) or the purified NisB protein together with purified His-tagged prenisin as substrate. Some of the cofactors tested during the enzyme assays were pyridoxal 5'-phosphate (P5P), adenosine monophosphate (AMP), adenosine triphosphate (ATP), and MnCl2. The reactions were performed at 30°C. We used reverse-phase high performance liquid chromatography (RP-HPLC) and electrospray mass spectrometry (ES-MS), or liquid chromatography–mass spectrometry (LC-MS) for the detection and analysis of the prenisin peptide products for the presence of any dehydro-residues. The preliminary enzyme assay results indicated that under the assay conditions used, we were unable to observe the dehydration. Although the in vitro reconstitution of some active LanM enzymes was successful, the in vitro reconstitution of the nisin dehydratase NisB or indeed other LanB enzymes is yet to be achieved despite significant efforts (Field et al. 2010). On the other hand, the overexpression of the nisB gene in wild-type nisin- and nisin variants-producing strains resulted in the complete conversion of Ser33 to Dha33, demonstrating the involvement of NisB in the dehydration process (Karakas Sen et al. 1999). Kuipers et al. (2004) demonstrated that in L. lactis strain expressing the nisBT genes, NisBT was sufficient to dehydrate and export the dehydrated non-lantibiotic peptides and the dehydrated prenisin. In another study, L. lactis strain expressing the nisBTC genes dehydrated and secreted a wide range of medically relevant non-lantibiotic peptides (Kluskens et al. 2005). More recently Shi et al. (2011) were able to show in vivo modification only in the presence of both NisB and NisC. It is therefore likely that NisB will only function in vitro in the presence of other maturation-specific proteins, implying that lantibiotic biosynthesis enzymes can be used for the creation of a wide range of peptides with novel bioactivities and of biostable analogues. It is also expected that the manipulation of modification enzymes will increase substrate tolerance and thus provide a more flexible approach for the introduction of lantibiotic-associated structures into non-lantibiotic peptides/proteins (Field et al. 2010).

In conclusion, the nisin biosynthesis gene nisB from L. lactis was successfully cloned into an expression vector and expressed in E. coli with an N-terminal extension of six histidine residues (His-tag). The cloning method described here has the advantage of avoiding a number of tedious cloning steps. His6-NisB protein was successfully purified in one step by Ni2+ affinity chromatography. The purification was carried out under native conditions in order to retain the protein activity and yielded a >90% pure product. His6-NisB protein could be obtained in milligram quantities using this procedure for use in in vitro structural and functional studies.